SEED PHYSIOLOGY, PRODUCTION & TECHNOLOGY Variation in Water-Use Efficiency and Its Relation to Carbon Isotope Ratio in Cotton Yehoshua Saranga,* Igal Flash, and Dan Yakir ABSTRACT Cotton (Gossypium spp.) is often exposed to drought, which ad- versely affects both yield and quality. Improved water-use efficiency (WUE = total dry malter produced or yield harvested / water used) is expected to reduce these adverse effects. Genetic variability in WUE and its association with photosynthetic rate and carbon isotope ratio (13C/12C) in cotton are reportedin this paper. WUE of six cotton cuitivars--G, hirsutum L., G. barbadense L., and an interspecific FI hybrid (G. hirsutum × G. barbadense, ISH), was examined under two irrigation regimes in two field trials. The greatest WUE was obtained by two G. hirsutum cultivars (2.55 g dry matter or 1.12 g seed-cotton L -1 H20); the ISH obtained similar or somewhat lower values, and two G. barbadense cultivars and one G. hirsutum cultivar exhibited the lowest values (2.1 g dry matter or 0.8 to 0.85 g seed- cotton L -1H20). Theseresults indicate that different cotton cultivars may have evolved different environmental adaptations that affect their WUE. Photosynthetic rate was correlated with WUE in only a few cases, emphasizing the limitation of this parameteras a basis for estimating crop WUE. Under both trials, WUE was positively correlated with carbonisotope ratio, indicating the potential of this technique as a selection criterion for improving cotton WUE. W; TER-USE EFFICIENCY is a key factor determining lant productivity under limited water supply. In physiological terms, WUE is defined as the ratio be- tween the rate of carbon fixed and the rate of water transpired by the leaf. In agronomic terms, however, it is defined as the ratio between total dry matter (DM) produced (or yield harvested) and water used (or plied). Agronomic WUE of crops has been improved significantly over the last half century, mainly by the introduction of modern irrigation and management techniques (Stanhill, 1991). Further improvement WUE requires exploitation of the physiological and ge- netic potential of the crop plants. Cotton (G. hirsutum and G. barbadense) is usually grown during the summerin arid and semiarid regions where water availability is often limited. Regardless of whether it is irrigated or not, cotton is often exposed to drought, which adversely affects both yield and lint quality. Improving the WUE of cotton plants could re- duce the adverse effects of drought and also improve productivity under non-stressful conditions. Modern cotton cultivars have been selected for specific proper- ties such as high yield, good fiber characteristics, and Y. Saranga and I. Flash, The Hebrew Univ. of Jerusalem, Faculty of Agricultural, Food and Environmental Quality Sciences, Dep. of Field Crops Vegetables and Genetics, P.O. Box 12, Rehovot 76100, Israel; D. Yakir, Weizmann Institute of Science, Dep. of Environmental Sciences and Energy Research, P.O. Box 26, Rehovot 76100, Israel. Received 14 April 1997. *Corresponding author (Saranga@agri. huji.ac.il). Published in Crop Sci. 38:782-787 (1998). adaptation to mechanical harvesting (Rosenow et al., 1983). In this process, the genetic potential available within modern cultivars for improving physiological traits, such as WUE, has been severely depleted. Never- theless, genetic variability in productivity and photosyn- thetic capacity still exists within cultivated cotton spe- cies. Pettigrew and Meredith (1994) reported genetic variation within G. hirsutum strains in photosynthetic rate, but there was little variation (in one out of 3 yr) in WUE,based on instantaneous gas-exchange mea- surements. High yield in G. barbadense was related to enhanced photosynthetic rate (Cornish et al., 1991) and to heat avoidance via greater evaporative cooling (Radin et al., 1994). However, it is unknown how this physiological variation affects seasonal WUE. A reliable evaluation of WUE is complicated and difficult to obtain. Physiological WUE can be evaluated by measuring CO2fixation and transpiration rates. These measurements are usually performed on a single leaf over a limited period of time (seconds or a few minutes at most) and, therefore, do not represent WUE of the entire plant or of the canopythroughout an entire day or season. On the other hand, agronomic WUE is based on measurements of DM and water consumption that are difficult to obtain. The development of an indi- rect, time-integrated selection criterion for WUE is, therefore, essential. Farquhar et al. (1982b) proposed that the variation in leaf carbon isotope ratio (13C/12C expressed with dif- ferential notation as g13C) in C3 plants depends on the ratio between intercellular CO2 concentration (Ci) and that of the ambient air (Ca). This relationship, based on theoretical considerations, has been demonstrated experimentally (Farquhar et al., 1982a; Downton et al., 1985; Seemannand Critchley, 1985). The C~/Ca ratio, which controls ~3C discrimination, can also be related to WUE,and high WUE is associated with higher (Ehleringer, 1989). Experimental evidence of the corre- lation between WUE and ~3C has been provided for several crops (Ehleringer et al., 1993). This approach provides a powerful tool for estimating integrated, long- term WUE in breeding programs. A major limitation of the carbon isotope approach is the fact that ~13C is influenced by any change in Ci. For example, both low stomatal conductance and high photosynthetic capacity will decrease Ci; and, in both cases, WUE will increase and ~13C will become more positive. It is essential, how- ever, to distinguish between these two scenarios because DM production increases only in the second case. When the goal is improvement of crop plants, it is also essential Abbreviations: DM, dry matter; HI, harvest index; ISH, interspecific hybrid; WUE, water-use efficiency. 782 Published May, 1998